Incremental Ideas

Travels in June to Tanzania and South Africa – and specifically to the TED Global conference in Arusha, and a World Economic Forum meeting in Cape Town – have had me thinking more directly about economic development than I have for the past few years. Specifically, I came out of these conferences amazed by the success of the mobile phone industry in Africa and wondering whether what lessons other sectors could take from the mobile phone industry.

I posited the idea of “incremental infrastructure” as a way to explain the success of the mobile phone industry – the idea that mobile companies were successful because they required modest up-front investment, generated revenue from a small, “atomic” investment and were subsidized in no small part by their users.

Evidently the idea has some legs. Gareth Cook at the Ideas section of the Boston Globe asked me if I would expand the idea for the newspaper – that expansion appears in today’s Ideas section. Gareth did a great job of forcing me to think about the big picture implications of the idea and the possibility that incrementalism could have implications for roads, air travel and other aspects of infrastructure, not just telecoms and power.

Writing for the newspaper – a form in which length is neccesarily constrained – has made me realize that I want to think and write at more length about this idea. The more I look into African approaches to infrastructure, the more I’m starting to think that incrementalism is the norm, not the exception. In failed states, there’s no opportunity for governments to make major infrastructure investments – inasmuch as systems exist, they’ve been built incrementally through modest investments by local entrepreneurs. Airlines in the Democratic Republic Congo are a classic example – they serve the function that a road or railroad system would serve in a luckier nation, and they are provided by “airlines” that frequently hasve assets that are no more than a single airplane. It’s not the transport infrastructure you’d wish for a nation, but it’s better than nothing, and enables commerce that couldn’t happen otherwise.

My friend and colleague Mike Best has challenged me on the idea of whether the examples I’m offering are really “micro” enough. By that definition, he’s most interested in “picoinfrastructure”, ways that small communities can build data infrastructure for a few thousands of dollars, not the single-digit millions I’ve pointed to in my examples from the mobile phone business. The examples Mike is offering mostly outline a model of “self-provisioning”, one where you’re building a network to meet your own needs because no one else has built that infrastructure. Where self-provisioning meets incremental infrastructure, I think, is where you overbuild for your personal needs with the goal of selling that capacity to your neighbors. Mining and oil companies self-provision infrastructure all over the African continent. Visit a gold mining center in Ghana and you’ll find infrastructure that outpaces that in many secondary cities. But that infrastructure isn’t shared and resold – it might be vastly more cost-effective if it were, and would likely address some community concerns about the impact of extractive industries. That said, I can imagine incremental infrastructure being built at mini, micro and pico-levels. A farmer investing in water pumping equipment that could irrigate both his fields and neighboring fields might be builting pico-level incremental infrastructure, while a mobile phone company that built power plants to provide energy to mobile phone base stations, and used excess capacity to run irrigation pumps might be working on a micro or mini scale.

Presenting the idea of incremental infrastructure to doctoral students at a summer session at Harvard raised another set of questions. The European students in the crowd were deeply suspicious of the market focus of incrementalism, and the idea that the government’s role in these proposals was, basically, to get out of the way. One response to this suspicion is to point to the poor record many African nations have had in building large-scale infrastructure, a fact that the African students in the crowd were happy to echo. Another is to suggest that the role for government is to actively get out of the way. The Investment Climate Facility for Africa is beginning to research constraints on operating businesses in African nations and recommend process reforms which make it easier for businesses to be founded and become operational. Governments that wish to encourage incrementalism can begin following these suggestions and changing their business environments.

As Mike suggested in this discussion, the most productive thing governments can do in this space is regulate intelligently. The response to air safety in the DRC is not to dismiss the incremental solution that’s been built up and attempt to replace it with a sole national carrier – it’s to regulate airlines and ensure that anyone flying is meeting minimum safety standards. Intelligent regulation can also prevent the “specter” of multinational dominance the students were worried about – Congo didn’t need to permit Alieu Conteh’s CWN and Vodacom Congo to merge. Had they been concerned about overconsolidation, they could have blocked the merger to encourage competition, or offered more licenses to encourage a third firm to expand in the nation.

Governments and aid agencies might also be able to assist with incrementalist strategies by focusing on remittance. Remittances sometimes create infrastructure on a highly local level – a water pump or generator for a single family, generally. Mexican communities have been experimenting with matching programs that will contribute public dollars or aid dollars to community projects funded via remittance – a worker might choose to send $150 to his family and $50 to a community school fund, especially if he knew the $50 would be matched 3 to 1 to build a school for his children. Matching of remittance for incremental projects has a very different “feel” from taxing remittance – instead of supporting the entire government infrastructure, the monies collected (voluntarily, I’d suggest) are guaranteed to focus in the area a worker wants to see benefit. They’re broader than the familial benefits of traditional remittance, but smaller than the national benefits generated by taxing remittances.

I’m a little wary of overfocusing on the incremental infrastructure idea. I’ve watched my friend Jamais Cascio see his idea of the ecological footprint of a cheeseburger spiral somewhat out of control, to the point where he’s now posted about the ecological footprint of making a TV show to discuss his research on the cheeseburger meme. (How’s that for meta?) But I think there’s a train of thought here worth exploring, and I hope to write something more substantive on this topic later this year. In the meantime, I’d love your reactions to the Globe piece and particularly your ideas for examples of how incrementalism is already working in developing nations, ways in which problems could be solved via incremental approaches, and problems inherent in the incremental infrastructure approach. Think of it as an incremental strategy for making this idea a more robust one. :-)

This entry was posted in Africa, Developing world. Bookmark the permalink.

11 Responses to Incremental Ideas

  1. Pingback: mtl3p » Blog Archive » summer

  2. June Arunga says:

    Hi Ethan,
    We met at TED Global in Arusha. I really enjoyed your Globe piece. I made a similar argument on a BBC News Night Piece in Feb titled “The Cell Phone Revolution in Kenya”, which was based on a paper I had written earlier that year, published by Istituto Bruno Leoni, Milan Italy. I like your term “incremental-infrastructure” and the non-polemic tone in your article and blog piece. Mine focused more on making a case for the private sector model of financing and execution of infrastructure provision, with lessons learned from the Mobile phone industry’s experience in Kenya. The reality is that while the taxation for various services goes on, most people pay for the actual service delivery from private providers , who would fit the “incremental-infrastructure” description. Take water provision in African cities, mostly done by everything from individuals carrying it on their heads for sale door to door to those who use pick up truck to ferry it neighborhood to neighborhood, to others who have several wells and tankers that they use to supply homes that have the money to build their own reservoir tanks. Another example would be in the provision of security services, where state provision has failed (barbed wire, welders who make window and door grills for sale, individuals who offer their personal services as guards, small companies that offer very cheap security guard services to home and businesses, to state of the art security companies, all competing in the same market), lighting, education, and almost everything else in mot African countries are de-facto privately provided. One of the main differences between these other industries and mobile telephony, is that the later was sanctioned by law in many countries, and could therefore receive larger capital investment, technological know-how and could thus be scaled, while in most cases water, electricity, education and security etc are still considered in many places are the government’s responsibility and so not accorded the same ease of licensing. I would encourage you to go further in exploring your idea. Without meaning to be dramatic… ultimately for many people around the world, this is a matter of life and death. Lack of clean water that has to be filtered, treated and pumped to homes, is the leading cause of death through water borne disease in several countries, so is lack adequate health care facilities that require electricity for refrigeration of medicines and vaccines and surgical procedures, and the list goes on and on. The implications of finding a working scalable model for infrastructure provision are huge. Pardon the ramble and any half baked thoughts expressed above. Point is, I so see what you see, think your labeling of it is useful and your framing of your arguments engaging even to those who have ideological reservations about markets.

  3. Pingback: ThaRum - Musings from Cambodia

  4. Pingback: The Status Quo and Radical Ideas | White African

  5. Ethan, I learned of your great article (and term incremental infrastructure) from David Weinberger’s post. This comes at a good time for me as I am working on two interrelated ideas in this regard. I’m putting together a proposal for software for marginal Internet access so that people in places like Africa could at their Internet cafe, which might be a 5 mile walk away and cost $3 an hour, download online activity they care about (letters, wiki, chat) onto their flash stick, go home to their offline computer and read and respond, then come back a week later and upload their responses. This would open the way for incremental rollout of local wi-fi networks – the knowledge needed and the online assistance needed could be there via the flashstick connection so that they could develop the skills to set up their own local communications network at their village. Say it covers 1 km across their village and does not link up to the global Internet yet I believe as Ethan shows that there is real local value. Indeed, the economic value of phones is first for local phone calls! Then you can keep adding wireless access points (100 USD) or used computers (200 USD) until finally in two or three years you can hook up to the global Internet. Which is much more likely than the other way around waiting for the global Internet to come to you. So encouraging the creation of local networks – which can have local value but also allow people to participate in global knowledge work that can be done mostly offline – imagine people contributing to Wikipedia articles etc. I appreciate more links to such ideas and here is our proposal that I’m writing: I wonder if you know who might see the business value and be able to fund $25,000 of work that we might get this rolling. I invite you and all to our lab’s chat Thursday, August 9, 2007 at 2:30 pm London time, 9:30 am New York time on this subject. We have excellent attendance from Africa, see our archives: and our letters at I’m also active at Rising Voices.



  7. Pingback: Global Voices Online » Bloggers Discuss Africa's Informal Economies

  8. Andrew Mack says:

    Incremental or not, what Africa needs is Entrepreneurial Infrastructure

    By Andrew Mack

    In his piece last month about “incremental infrastructure”, Ethan Zuckerman makes a number of excellent points about the recent development of infrastructure in Africa. Using his example of the entrepreneur who put up cell towers in the Democratic Republic of the Congo, he rightly observes that there are opportunities to think beyond the traditional, top-down structures of infrastructural development.

    He cites the logistical and budgetary problems of many nations as they seek to build out not just the famous “last mile”, but in cases like DRC, many of the basic earlier miles that need to be in place if a country wants to be connected – by road, by power grid, or by wireless. And, while he doesn’t dwell on one of the real reasons for this failure – Government disorganization or outright corruption – he hints at it as a driving force which creates both the space and the need for other approaches.

    However, while the idea of incremental infrastructure is interesting, I would argue that at least to some extent, Ethan’s argument misses the larger point. It is not incremental infrastructure so much as “entrepreneurial infrastructure” that Africa needs and has shown it wants.

    By focusing on the example of cell towers in DRC, Ethan may have chosen the one item best suited for incremental infrastructure. But consider roads… if a community – or a firm – decides to build an incremental piece of road, and who pays? Who maintains the road? Who sets the safety standards (as road accidents are an epidemic in many African countries today)? And what if the road doesn’t connect in to a larger grid? Clearly, while cellphones may not need coordination to function, most other pieces of infrastructure – roads, energy, etc. – do. And they need standards.

    Moreover, there are issues of economy of scale and policy. Consider the case of neighboring Uganda. In Uganda, cellphone licenses were bid out, encouraging competition, and cellphone use has grown from some 5,000 lines in 1998 to more than 2.6 MILLION today. The major carriers have invested – and made – millions, and I would argue, have done more for Uganda’s development than most of the major donors over the last decade. A licensing regime that favored incremental providers might have brought service to a few villages, but today, CelTel is the largest taxpayer in the country, serving the entire nation and recently, offering no-roaming service across the sub-region – in Kenya, Tanzania, and recently also in DRC. An incremental approach would not have been able to provide this service, pure and simple.

    What we need to do is re-orient our thinking, I believe. Rather than focusing on the challenges – and there are many – we need to step out of our past frame and see the markets as what they are: big and underserved. What we need is not so much small (incremental) infrastructure as infrastructure that is constructed by people with an entrepreneurial mindset. If the Government of Kenya is prepared to invest in wiring classrooms and has both the scale and technical savvy to pull it off, that’s great. If the private sector can do it better, then the Government should act as facilitator. In some instances, a public-private approach will be the best.

    Without question, Ethan makes a good point about the effectiveness of many large projects in Africa, especially famous dam projects and the like. Still, this is not unique either to Africa, or to energy. Corruption and a lack of oversight will ruin a project, whether it’s managed by a poor African Government or Halliburton. Especially in infrastructure, the key is getting value for money. Everybody – Governments, the private sector, and consumers themselves – all need to think entrepreneurially.

    In the end, while small may be beautiful in many things, I wouldn’t want my own water system in Washington, DC any more than my friends in Lamu would want their own. What they want is a water system that works. Based on my observations from more than 20 years work on the continent, I would argue that a focus on the incremental could – while providing solutions in some areas – actually hurt efforts to build a more complete and more robust African infrastructure with the policies and investments to make it sustainable.

    While there will always be underserved areas where other options might not be possible, incremental infrastructure would be a poor substitute for the kind of top shelf, state of the art infrastructure Africans are looking for, the kind of infrastructure that could help them compete in the global economy.

    Andrew Mack is the Founder and Principal of AMGlobal Consulting, and a former World Bank official. He can be reached at:

  9. Pingback: Africa’s Infrastructure-Incremental or Entrepreneurial? «

  10. Pingback: AmGlobal » Blog Archive » Entrepreneurial Infrastructure

  11. Pingback: Entrepreneurial Infrastructure - AMGlobal Consulting

Comments are closed.